About

Senior Platform & Performance Engineer (SaaS ERP) Goal: Own performance, search/analytics, and delivery pipelines for SaaS ERP. Ship the stack below, cut p95 latencies, and make reporting/search feel instant. What you’ll deliver (first 60–90 days) • ReadySet in front of Postgres for hot read paths (list/filters/aggregates). • Meilisearch for instant inventory search + autocomplete. • ClickHouse + ClickPipes (Postgres CDC) powering sub-second tenant dashboards. • ThoughtSpot Embedded liveboards + SSO wired into Caratiq. • Harness CI/CD pipelines with canary/rollback & policy gates. • JMeter load test suites + baseline SLAs and perf budgets. • Kong (Konnect or OSS) as API gateway (auth, rate limits, mTLS, keys). Core responsibilities • Optimize backend queries (EXPLAIN/ANALYZE, indexing/partitioning, query rewrites; tools like pganalyze/EverSQL/pgMustard). • Frontend performance tuning (React/TS): bundle analysis, memoization, virtualization, Core Web Vitals. • Speed up bulk jobs (imports/exports), move heavy side-effects to async workers, and add robust queues. • Define and track perf KPIs (p50/p95 API latency, throughput, error budgets); build dashboards/alerts. • Harden API surface via Kong (OIDC/OAuth2, scopes, rate limiting, WAF rules). Required experience (must-haves) • 6–10+ years total engineering; 4+ years backend on high-traffic web apps. • Postgres performance: indexing, query plans, connection pooling, read/write patterns. • ClickHouse or equivalent OLAP (e.g., Druid/BigQuery/Snowflake) with CDC/ELT experience. • Search engines: Meilisearch or Elasticsearch/OpenSearch in production. • CI/CD: built and owned pipelines (Harness, GitHub Actions, GitLab CI, or similar). • API gateways: Kong/Apigee/NGINX (policies, plugins, auth). • Load testing: JMeter or k6; can script realistic scenarios and interpret results. • Cloud & services: AWS (S3, IAM, networking), Docker; strong Linux fundamentals. • JavaScript/TypeScript proficiency (Node/NestJS; React basics for perf fixes). Nice to have • ThoughtSpot Embedded (or Metabase/Looker/Power BI Embedded) with SSO & multi-tenant embedding. • ReadySet/Materialize/CDC-based result caching; Debezium/Airbyte/ClickPipes. • Redis (queues, caching), GraphQL (Apollo) performance, Strapi v4 familiarity. • Observability: Datadog (APM, RUM, traces) Tech you’ll touch here Node/NestJS, React/TypeScript, Postgres, Redis, ReadySet, Meilisearch, ClickHouse + ClickPipes, ThoughtSpot Embedded, Kong, Harness, JMeter, S3, SQS/BullMQ, Datadog How we’ll measure success • -50% p95 on top 10 API endpoints; -75% time-to-interactive on key pages. • Less than 1s p50 query times for core product/search views; sub-second dashboard tiles. • Zero-touch deployments with canary/auto-rollback and green test gates. • Repeatable load tests + weekly perf report with clear owner & next actions.